Introduction Strategic Objectives of IT Operation for 2008 Census Constraints Conclusion

Transcription

1 For Expert Group Meeting Opportunities and advantages of enhanced collaboration on statistical information management in Asia and the Pacific, June 2011, Bangkok, Thailand Introduction Strategic Objectives of IT Operation for 2008 Census Constraints Conclusion

2 The NIS, which is the part of the Ministry of Planning, is focal point on statistical matters in Cambodia. The mandate of NIS is data collection, compiling and dissemination though the survey and censuses. The statistical units and sections contains the line Ministries and Provincial Planning Department. IT Operation in NIS has started since 1993 up to date with support of many development partners. The IT methods used in 1998 census and 2008 Census are similar methods using micro computer for data entry, computer editing, tabulation and dissemination products. IT Strategy for 2008 Census comprises a large array of activities, ranging from pre-census and post-census. Data Processing Plan Tabulation and Data Analysis The Dissemination Plan

3 Pre-census data processing activities Prepare Enumeration Area Maps and Databases Data processing site Procure the necessary equipment Training on Data Entry, Computer Editing, and Tabulation application development To develop the required training manuals, coding schemes, and software application. Post-census data processing activities: Check-in of census documents Manual coding and editing Data Entry Computer editing Tabulation of census results Release of Census figures in two stages: 1. Provisional Totals in Sep and 2. Final Result in Sept Forms of dissemination Population database Examples of macro-level Population Databases Data Storage More than 2,8 million questionnaires received at NIS Census Office. Questionnaires stored EA-wise in folders (approx. 120 questionnaire per EA). Each folder labeled with a unique bar code. To track movement of EA batches and to register against Census Frame. Storage onto shelving racks.

4 To prepare census questionnaires for the next stage, i.e. Data Entry. Consistency of information contained in each questionnaire to be verified/edited. Some four variables require coding. For example: Occupation farmer policeman Occupation farmer policeman Time consuming activity. Information on some 13.4m persons and 2,8 m households needs to be captured! 140 microcomputers and keyboard operators have been used. Data entry software is CSPro. A userfriendly package developed by US Census Bureau. Some 20% verification to minimize typing errors. Took 9 months for Data Entry

5 Errors in census data can be introduced at various stages: Respondent s s error Interview error Coding/editing error Data entry (typing) error During computer editing a set of computer programs is used to detect and correct above errors. Software is the Batch Editing module of CSPro. Results of computer editing are clean data sets, free of errors and ready for tabulation. In 1998 Census, 57 basic Priority Tables down to district level 11,513 priority tables have been produced. In 2008 Census, 70 basic Priority Tables and down to commune level 73,151 priority tables have been produced Census has also produced more for supplement tables. The analysis of census data by NIS is mainly confined to priority tables, chart and additional tables that may be produced for analysis.

6 Some 70 priority tables have been defined. Ad hoc tabulation will be produced upon request. Priority tables to be produced at national, provincial and district level. Limited tabulation can be performed on commune and village level. Software is the CENTS module of IMPS. Provisional census results, consisting of population figures by sex, were released on September 2008 presided by Deputy Prime Minister of RGC. Detailed census tabulation at national level was released after completion of Data Entry and computer editing, i.e. third quarter of September 7, 2009 by Samdech Akka Moha Sena Padei Techo HUN SEN, Prime Minister of the Kingdom of Cambodia.

7 Publication of printed tables and reports for general distribution. Dissemination on computer media for some users. On-line dissemination; Internet. Workshop/Seminar The Table Retrieval System stores the large number of census tables onto a single CD ROM and facilitates easy retrieval of selected tables. A user-friendly interface will be available to select multiple tables for multiple geographical areas. The application will also include a table viewer that allows exporting the tables or table cells to Excel spreadsheets for further manipulation and analysis

8 This dissemination product is based on an indicator database consisting of aggregated counts, rates and ratios for all possible geographical levels. A variety of indicators may be considered, such as: distribution by age groups and sex, household types, median age, dependency ratio, singulate mean age at marriage, literacy rate, employment rate, educational attainment, proportions of migrants, etc. The application allows to select areas and will include functionality to present some of the key indicators in a graphical format and to export the profile to Excel format. Greatly expands the usability and enhances the dissemination of census data. Database for micro-data (individual records) permit retrieval of data at any level of detail. Database for macro-data (aggregated data). Less versatile because predefined scope and formats.

9 Table-oriented databases. Time-series and indicators databases in which census results are integrated with other data sets, e.g. previous censuses and other inquiries such as health, education, agriculture statistics. Graphing and mapping databases. Mapping capabilities greatly improves the effectiveness of census information dissemination. Mapping and graphing databases also greatly improve the effectiveness of census dissemination as trends and patterns of the larger area, and distinct boundary and characteristics are more easily detected when displayed on maps. The project will make available map layers for the Country, Provinces, Districts, Commune and, if possible, Villages. Statistical databases consisting of aggregated count, ratios and rates for all geographical levels will complement these map layers. The layers will be in ESRI shape file and MapInfo table file formats. Data users are expected to acquire a copy of their preferred GIS software

10 CamInfo is the national adaptation of Dev Info software, a global initiative funded by the United Nations. Cam Info is Cambodia's Socio-Economic and demographic Indicator Database, providing a one-stop user-friendly computer program for storage, retrieval, comparison and dissemination of a wide range of indicators from different sources, including national surveys, censuses and administrative systems. CamInfo also allows the user to retrieve and compare indicator data values across multiple time periods, geographic levels, and other sub-group dis-aggregations. Data presentation is possible with tables, graphs and maps. 1. CD-ROM 2. External Hard Disk 3. Server

11 Paucity of supervisory IT staff in the Census Data Processing Division. Sometimes virus-affected computers. Coding wrong Inconsistency of data Administrative area changes Suitable incentive to Data Processing staff may be needed The IT strategy for 2008 General Population Census of Cambodia played an important role and served as an essential tool to successfully complete the census work. The IT strategy covered the pre-census and postcensus operations. The pre-census activities included check-in phase, site for data processing, and training. The post-census activities covered data entry, computer editing, tabulation and dissemination operations. It would be better if lessons are drawn from the constraints mentioned above to improve the IT in the future censuses

12